Why Most Data Projects Fail & How to Avoid It • Jesse Anderson • GOTO 2023

Learn why most data projects fail and how to avoid common pitfalls by understanding the roles and responsibilities of data scientists, engineers, and operations specialists, and how clear planning and execution can lead to success.

Key takeaways
  • Many data projects fail due to various reasons, including lack of business value, technological complexity, and poor organizational structure.
  • To succeed, projects must demonstrate value and feasibility within a reasonable timeframe.
  • Data scientists are not responsible for operations and engineering, but rather produce data products for consumption by others.
  • Data engineers focus on software engineering and operations, ensuring data products are scalable and maintainable.
  • Operations engineers focus on operationalizing data frameworks, ensuring they are repeatable and manageable.
  • Clear definitions are crucial for understanding the roles and responsibilities of data scientists, engineers, and operations specialists.
  • Projects should be planned with a clear definition of what needs to be done, how it will be done, and who will be responsible for it.
  • Data engineers are essential for data product success, but their skills are often underappreciated or overlooked.
  • Data projects often fail due to a lack of execution, rather than technological issues.
  • The right people, with the right skills and expertise, are essential for project success.
  • Projects should be designed to deliver tangible business value, rather than relying on technical complexity.
  • Data scientists are not expected to be experts in operations and engineering, but rather focus on producing data products.
  • Data engineers are responsible for making data products scalable, maintainable, and efficient.
  • Operations engineers ensure data frameworks are operationalized and repeatable.
  • Clear planning and execution are crucial for data project success.
  • Projects should focus on delivering business value, not just technical complexity.
  • The right ratio of data scientists, engineers, and operations specialists is crucial for success.